Detection of Circulating Tumor Cells and Circulating Tumor DNA Before

Breast Cancer Research and Treatment (2019) 177:447–455 https://doi.org/10.1007/s10549-019-05326-5

CLINICAL TRIAL

Detection of circulating tumor cells and circulating tumor DNA before and after mammographic breast compression in a cohort of breast cancer patients scheduled for neoadjuvant treatment

Daniel Förnvik1 · Kristina E. Aaltonen2 · Yilun Chen3 · Anthony M. George3 · Christian Bruefer3 · Robert Rigo3 · Niklas Loman3,4 · Lao H. Saal3 · Lisa Rydén5,6

Received: 25 April 2019 / Accepted: 17 June 2019 / Published online: 24 June 2019 © The Author(s) 2019

Abstract Purpose It is not known if mammographic breast compression of a primary tumor causes shedding of tumor cells into the circulatory system. Little is known about how the detection of circulating biomarkers such as circulating tumor cells (CTCs) or circulating tumor DNA (ctDNA) is afected by breast compression intervention. Methods CTCs and ctDNA were analyzed in blood samples collected before and after breast compression in 31 patients with primary breast cancer scheduled for neoadjuvant therapy. All patients had a central venous access to allow administration of intravenous neoadjuvant chemotherapy, which enabled blood collection from superior vena cava, draining the breasts, in addition to sampling from a peripheral vein. Results CTC and ctDNA positivity was seen in 26% and 65% of the patients, respectively. There was a signifcant increase of ctDNA after breast compression in central blood (p = 0.01), not observed in peripheral testing. No increase related with breast compression was observed for CTC. ctDNA positivity was associated with older age (p = 0.05), and ctDNA increase after breast compression was associated with high Ki67 proliferating tumors (p = 0.04). CTCs were more abundant in central compared to peripheral blood samples (p = 0.04). Conclusions There was no signifcant release of CTCs after mammographic breast compression but more CTCs were present in central compared to peripheral blood. No signifcant diference between central and peripheral levels of ctDNA was observed. The small average increase in ctDNA after breast compression is unlikely to be clinically relevant. The results give support for mammography as a safe procedure from the point of view of CTC and ctDNA shedding to the blood circulation. The results may have implications for the standardization of sampling procedures for circulating tumor markers.

Keywords Circulating tumor cells · Circulating tumor DNA · Breast compression · Breast cancer · Mammography · Neoadjuvant

Abbreviations Lao H. Saal and Lisa Rydén shared senior authorship. CTC Circulating tumor cells ctDNA Circulating tumor DNA Electronic supplementary material The online version of this RNA-seq RNA sequencing article (https://doi.org/10.1007/s10549-019-05326-5) contains supplementary material, which is available to authorized users.

* Daniel Förnvik 4 Department of Oncology, Skåne University Hospital, Lund, [email protected] Sweden 5 Department of Clinical Sciences Lund, Division of Surgery, 1 Department of Translational Medicine, Medical Radiation Lund University, Lund, Sweden Physics, Lund University, Malmö, Sweden 6 Department of Surgery and Gastroenterology, Skåne 2 Department of Laboratory Medicine, Division University Hospital, Malmö, Sweden of Translational Cancer Research, Lund University, Lund, Sweden 3 Department of Clinical Sciences Lund, Division of Oncology and Pathology, Lund University, Lund, Sweden

Vol.:(0123456789)1 3 448 Breast Cancer Research and Treatment (2019) 177:447–455

MAF Mutant allele frequency trapped in the capillary bed of the frst organ they encoun- TNBC Triple-negative cancers ter [27]. In breast cancer, an autopsy study by Peeters et al. [28] showed that CTCs were trapped in the lung microvas- culature in four of the nine patients who all had high CTC Introduction counts (> 100). Thus, it is likely that the number of CTCs found in the peripheral blood system is not representative Circulating tumor markers such as circulating tumor cells of a possible release of tumor cells from the primary tumor (CTCs) and circulating tumor DNA (ctDNA) can be found during manipulation such as breast compression during in the blood of cancer patients. As a liquid biopsy, these mammography or surgery. The diference in CTC number markers complement solid biopsies and have the advantage between central and peripheral blood is possibly even more of being physically more accessible and patient-friendly pronounced after specifc interventions compared to a more than traditional tissue biopsies. This provides a possibil- steady-state-like condition of metastatic disease [24, 26]. ity for prognosis prediction, closer monitoring of treatment To our knowledge, ctDNA levels have not been used to response and disease progression, identifcation of drug tar- study a possible release of tumor cells or tumor cell debris gets, as well as an opportunity for early detection of recur- after breast compression or any other mechanical interven- rence. The presence of CTCs in the blood of patients with tion in breast cancer. Relatively few studies have so far primary breast cancer has been shown to be an independ- compared the levels of both CTCs and ctDNA in the same ent predictor of decreased disease-free and overall survival clinical patient cohort at identical time points and our under- [1, 2], but the treatment predictive value of the cells is still standing of the relationship between the two liquid tumor under debate [3, 4]. The CTC methodology in primary breast markers is limited. However, both the level of ctDNA and cancer is also limited by the low number of detected cells, the number of CTCs have been shown to have a prognostic which makes enumeration and evaluation statistically chal- value in mainly metastatic breast cancer cohorts [29, 30]. lenging [5]. ctDNA, the cell-free DNA that originates from Mutation analysis of ctDNA and single CTCs suggests that cancer cells, is a promising biomarker whose prognostic and ctDNA refects the heterogeneity of mutations found in indi- treatment predictive power is emerging [6, 7]. Recent stud- vidual CTCs [30], but ctDNA levels have been found to have ies have shown that quantifcation of specifc mutations in a higher correlation with tumor burden than CTCs [29]. ctDNA can be associated with early detection of metastases The aim of this study was to investigate how the pres- and therapy resistance in breast cancer as well as in other ence of CTCs and ctDNA are afected by breast compres- diagnoses [8–12]. sion during mammography in patients with primary breast The risk that tumor cells are released into the bloodstream cancer. Special emphasis was made on comparing circulat- from a primary tumor during surgical interventions has been ing tumor marker burden between the central and peripheral addressed in a few studies [13–17], although how and when blood circulation. tumor cells are shed as well as the clinical importance of this release is poorly understood [18]. Animal studies have also found that physical manipulation of a primary tumor Materials and methods by applying pressure to it causes tumor cell dissemination [19–21]. We have previously investigated if mammographic Patient cohort and clinical parameters breast compression in patients with an already present breast tumor could cause shedding of tumor cells to the peripheral The patient cohort comprises preoperative patients within the circulation [22]. We found no indications that this would be ongoing SCAN-B trial (Clinical Trials ID NCT02306096) the case in a pilot study of 24 patients with primary breast at Lund University and Skåne University Hospital, Sweden cancer. [31, 32]. During 2015–2016, 31 patients scheduled for neo- However, the confguration of the human blood circula- adjuvant therapy volunteered to do an extra mammography tion can cause tumor cells released from the breast to pass after diagnosis and were included in the present study. The through the capillary vasculature of the lungs before reach- patient mean age was 51.9 years (range 33–74 years) and ing the peripheral blood vessels. In our previous study [22], the mean compressed breast thickness and applied com- CTCs captured only in the peripheral blood might have pression force during the examination were 55.8 mm (range resulted in an underestimation of CTC number. It has been 26.5–77.0 mm) and 103.4 N (range 71.5–123.1 N), respec- shown that a higher number of CTCs can be found in cen- tively, as indicated by the mammography system (Mammo- tral compared to peripheral venous blood in patients with mat Inspiration, Siemens Healthineers, Erlangen, Germany). metastatic breast cancer [23] as well as in other diagnoses All patients gave written informed consent and the study was [24–26]. Animal studies of colon carcinoma cells have approved by the Regional Ethical Review Board in Lund, shown that the majority (80–100%) of tumor cells could be Sweden (diary number 2014/521).

1 3 Breast Cancer Research and Treatment (2019) 177:447–455 449

Clinical data including biomarker expression, histologi- ctDNA analysis cal subtype, and nodal status were retrieved from pathology reports and the patient’s clinical charts. Information on bio- Candidate somatic mutations for ctDNA measurement were marker expression was based on analysis from the core nee- obtained from RNA sequencing (RNA-seq) data generated dle biopsy before initiation of neoadjuvant therapy. Estrogen within SCAN-B [31, 34]. Twenty-eight of the 31 patients receptor (ER) positivity was defned as ≥ 10% positive can- had available tumor RNA-seq data. Sequencing, base call- cer cells, human epidermal growth factor receptor 2 (HER2) ing, FASTQ fle processing, and fltering were performed positivity was defned by immunohistochemistry (IHC) or as previously described [34]. Using a Snakemake workfow, in situ hybridization (ISH) as (IHC3+) or ISH-positive cells, reads in FASTQ format were aligned to the human reference and Ki67 positivity was defned as > 20% positive cancer genome GRCh38.p8 (including alternative sequences and cells. Information from mammograms, ultrasound images, decoys, and patched with dbSNP Build 147) using HISAT2 and breast tomosynthesis was compiled into one measure 2.0.5 [35] (with default options except --rna-strandness RF, of tumor size. --rg-id ${ID_NAME}, --rg PL:illumina, --rg PU:${UNIT}, --rg SM:${SAMPLE}), and duplicate reads were marked using SAMBLASTER 0.1.24. Variants were called using Blood sampling VarDict-Java 1.5.0 [36] (with default options except -f 0.02, -N ${SAMPLE}, -b ${BAM_FILE}, -c 1, -S 2, -E 3, -g 4, All patients had a central venous access to allow admin- -Q 10, -r 2, -q 20), and annotated with dbSNP build 150 and istration of intravenous neoadjuvant chemotherapy, which COSMIC v84 using vcfanno 0.2.8 [37]. enabled blood collection from superior vena cava, draining From the RNA-seq mutation calling, one somatic muta- the breasts. A dedicated research nurse attended the patient tion for each patient was selected for IBSAFE assay design during the mammography examination and acquired blood for ultrasensitive mutation detection. IBSAFE© (SAGA samples before and after mammographic breast compres- Diagnostics AB, Lund, Sweden) is an enhanced droplet sion, frst from central venous access and secondly from a digital PCR technology with signifcantly improved sensi- peripheral vein at both occasions. The median time for blood tivity and specifcity, allowing for quantifcation of alleles sampling after compression was 2 min (range 0–5 min) for to 0.001% mutant allele frequency (MAF) [George et al. central blood samples and 7 min (range 5–23) for peripheral manuscript in preparation]. IBSAFE assays targeting a blood samples. At each time point, 10 ml whole blood was somatic mutation were designed for 20 patients and the collected in CellSave tubes (Menarini Silicon Biosystems, assays validated using 6 ng of corresponding tumor DNA Bologna, Italy) for CTC analysis and 10 ml whole blood was as positive control and 180 ng of human normal genomic collected in Cell-Free DNA Blood Collection Tubes (Streck DNA (Promega, Madison, USA) as negative control, con- Inc., Omaha, USA) for ctDNA analysis. The blood samples frming a lower limit of detection of at least 0.0017% MAF were transported at room temperature and subsequent analy- for each assay. ses were performed within 96 h after sample taking. Whole blood collected in Streck tubes were centrifuged at 2000×g for 15 min at room temperature to fractionate plasma, followed by clearing of the plasma fraction by cen- CTC analysis trifugation at 10,000×g for 15 min at 4 °C. Cell-free DNA was isolated using the QIAamp Circulating Nucleic Acid Kit The blood samples were analyzed for CTC number using or the QIAamp MinElute ccfDNA Midi Kit (Qiagen, Hilden, the FDA-approved CellSearch © system (Menarini Silicon Germany), of which 20% of the eluate used for IBSAFE Biosystems). Briefy, a ferrofuid-conjugated epithelial cell reactions and measurement of mutant and wild-type ctDNA adhesion molecule (EpCAM)-directed antibody was used copies and calculation of MAF. to separate CTCs from the majority of white blood cells. Fluorescent staining with DAPI (nuclear staining), cytokera- Statistical analysis tin (CK) 8, 18, 19-directed PE-conjugated antibodies, and CD45-directed APC-conjugated antibodies were applied Clinical and patient-specifc characteristics were compared to identify CTCs (DAPI +/CK+/CD45–). Two independ- between patients that had ≥ 1 CTC/≥ 0.01% MAF present ent and accredited technicians manually evaluated images in any sample and patients with 0 CTCs/0% MAF in all of CK + events selected automatically by the CellTracks II samples. Agreement between CTC- and ctDNA-positive system (Menarini Silicon Biosystems). The method has been patients was analyzed using Cohen’s kappa statistics. The described in detail elsewhere [33]. Cut-of for CTC posi- Mann–Whitney U-test was used to compare the distribution tivity was ≥ 1 CTC/7.5 ml blood as suggested by a recent of continuous variables. For categorical variables, Fisher’s review of primary breast cancer [1]. exact test was used in all comparisons due to less than fve

1 3 450 Breast Cancer Research and Treatment (2019) 177:447–455 expected cases in at least one of the groups in all cross- patients, without reaching statistical signifcance (Table 1). tables. Statistical analysis of all characteristics was also Notably, 4/4 triple-negative breast cancers (TNBC) and 4/4 performed between patients that had an increase in CTC T4 staged cancers were all ctDNA positive. number/% MAF after compression with patients that did not Thirty patients had CTC results from the central blood have an increase in CTC number/% MAF after compression. sample before and after breast compression and 22 of these A non-parametric Wilcoxon signed-rank test was applied to patients had 0 CTCs at both time points. Five of eight test for CTC diferences/% MAF changes between before patients with detectable CTCs had an increased number of and after compression. For comparison between central and CTCs after compression (p = 0.19) (Fig. 2a). The average peripheral CTC/ctDNA measurements, a sign test was used. CTC increase was 3.2 cells (median 1.0 cell). Only two eval- All statistical analyses were performed in IBM SPSS Sta- uable patients had detectable CTCs in the peripheral blood tistics (version 24, IBM, Armonk, NY, USA) and p values sample (Fig. 2b). Both central and peripheral % MAF gen- < 0.05 were considered signifcant. erally increased after compression with the latter reaching signifcance (p = 0.08 and p = 0.01) (Fig. 2c, d). The average increase of % MAF was relatively small, 0.77 and 0.35 Results (median 0.35 and 0.22% MAF) for central and peripheral, respectively. Of the 20 patients with assessable ctDNA sam- In total, 8/31 patients (26%) had ≥ 1 CTC in at least one ples before and after breast compression, eleven and eight of the blood samples taken before or after mammographic patients had 0% MAF in central and peripheral plasma sam- breast compression (Fig. 1). Correspondingly, 13/20 patients ples, respectively, at both time points. (65%) had ≥ 0.01% MAF and were defned as ctDNA posi- The median fraction of Ki67-positive cells was 66% tive. No agreement was found between CTC- and ctDNA- (range 30–90%) in the fve patients that had an increase in positive patients (κ = 0.02, p = 0.92) (A plot of CTC count CTC number after compression, compared to 45% (range versus % MAF can be found in supplementary Fig. S1). 15–95%) in patients with no increase (p = 0.31) (Supplemen- Patient and tumor characteristics of the whole cohort, as tary Table 1). Also, Ki67 fraction was signifcantly higher well as of CTC/ctDNA-positive and CTC/ctDNA-negative in the group with increasing ctDNA after compression, 45% patients separately, are shown in Table 1. No patient or path- versus 30% (p = 0.04) (Supplementary Table 2). No other ologic characteristics were statistically associated with CTC factors were diferentially expressed between patients with positivity. Larger tumor size, non-ductal histological sub- an increase in CTCs/ctDNA levels and patients with a sta- type, and older age were more predominant in the CTC-pos- ble or a decrease in CTCs/ctDNA levels after compression. itive group but the diference was not statistically signifcant. However, the histological type of the primary tumor seemed ctDNA positivity was associated with higher age (p = 0.05). to difer between patients with an increase in the number of Higher Ki67, ductal histological type, and triple-negative CTCs and patients with an increase in the levels of ctDNA breast cancer were more predominant in ctDNA-positive after compression (Supplementary Tables 1 and 2).

Fig. 1 Examples of CTCs detected with the CellSearch system from a patient in the study

1 3 Breast Cancer Research and Treatment (2019) 177:447–455 451

Table 1 Comparison of patient and tumor characteristics between patients positive for CTCs and ctDNA (≥ 1 CTC/≥ 0.01% MAF) and patients with no CTCs/% MAF Total (N = 31) CTC negative CTC positive (N = 8) p value Total (N = 20) ctDNA ctDNA p value (N = 23) negative positive (N = 7) (N = 13)

Age (years) Median (range) 50 (33–74) 47 (33–74) 56 (43–71) 0.21a 51 (35–74) 46 (35–62) 58 (40–74) 0.05a < 50 16 13 3 0.43b 10 5 5 0.35b ≥ 50 15 10 5 10 2 8 Tumor size and stage Median size, mm 30 (4–90) 30 (4–90) 38 (16–80) 0.21a 30 (4–80) 24 (8–80) 30 (4–80) 0.60a (range) T1 (< 20 mm) 8 7 1 0.64b 6 3 3 0.61b T2–T4 (20 mm or 23 16 7 14 4 10 higher) Nodal stage N0 4 3 1 1.0b 2 0 2 0.52b N+ 27 20 7 18 7 11 ER Negative (10% or 8 6 2 1.0b 4 0 4 0.25b lower) Positive (> 10%) 23 17 6 16 7 9 HER2 Negative 25 18 7 1.0b 17 5 12 0.27b Positive 6 5 1 3 2 1 Ki67 Median % of cells 45 (15–95) 45 (20–95) 49 (15–90) 0.61a 40 (15–90) 30 (15–90) 45 (20–90) 0.19a stained (range) Low (20% or 3 2 1 1.0b 3 1 2 1.0b lower) High (> 20%) 28 21 7 17 6 11 Breast cancer subtype ER+ 18 12 6 0.63b 13 5 8 0.20b HER2+ 6 5 1 3 2 1 TNBC 7 6 1 4 0 4 Multifocality No 22 17 5 0.64b 15 4 11 0.29b Yes 8 5 3 5 3 2 Missing 1 1 Histological subtype Ductal 23 19 4 0.15b 13 3 10 0.17b Other 8 4 4 7 4 3 Detection mode Screening 9 7 2 1.0b 8 2 6 0.64b Symptomatic 22 16 6 12 5 7 a Mann–Whitney U-test b Fisher’s exact test

CTCs were more abundant in central compared to central and peripheral sampling (8/20 favoring higher peripheral blood in 8/10 positive samples (p = 0.04) % MAF in the central blood sample, p = 0.50) (Fig. 3b). (Fig. 3a). Forty-nine comparisons between central and Twenty comparisons between central and peripheral blood peripheral blood contained 0 CTCs in both samples. There contained 0% MAF in both samples. was no signifcant diference in % MAF levels between

1 3 452 Breast Cancer Research and Treatment (2019) 177:447–455

A B

C D

Fig. 2 The number of CTCs found before and after mammographic and peripheral (d) plasma, where two patients are represented by a breast compression in central venous access (a), where two patients line going from 0 to approximately 0.04 and from approximately 0.03 are represented by a line going from 0 to 1 CTC and from 1 to 0 to 0, respectively. Patients with an increasing value are plotted with CTCs, respectively. The corresponding number of CTCs before and red lines, decreasing values in blue, and constant values in green. All after compression in peripheral blood (b). Figures for mutant allele patients that did not have any circulating tumor markers are summa- frequency before and after breast compression in central (c), where rized in one line at number/frequency = 0 two patients are represented by a line from 0 to approximately 0.35,

Discussion study, was associated with higher age (p = 0.05) and a trend was noted that a more aggressive tumor phenotype, includ- CTCs were detected in 26% of the patients before start of ing high Ki67, TNBC, and T4 staged cancers, favors ctDNA neoadjuvant therapy for primary breast cancer. This is in line positivity (not statistically signifcant). with a recent meta-analysis where 25.2% of breast cancer For CTCs, no increase was seen in either central or patients had CTCs before onset of neoadjuvant chemother- peripheral blood after mammographic breast compression apy (deemed independent of blood sampling volume) [2]. (p = 0.19 and p = 0.37, respectively). However, both central ctDNA was positive in 65% of the patients. The concordance and peripheral ctDNA levels increased after breast compres- between CTCs and ctDNA has been shown to be higher in sion (p = 0.08 and p = 0.01, respectively). Only one patient metastatic breast cancer patients [29] as compared to what had a CTC count diference of > 5 cells/7.5 ml between was found in this study, which is most likely due to the lower samples taken before and after compression (Fig. 2a). This rate of CTCs in primary breast cancer (Supplementary Fig. suggests a lack of a larger bolus release of cells during S1). ctDNA positivity, defined as ≥ 0.01% MAF in this breast compression of women with primary breast cancer.

1 3 Breast Cancer Research and Treatment (2019) 177:447–455 453

Fig. 3 CTC (a) and ctDNA (b) detection in central and peripheral samples, a higher mutant allele fraction was found in the peripheral sample pairs. In 8/10 CTC-positive pairwise samples, a higher num- plasma sample (p = 0.50). In pairwise samples 2, 4–7, no ctDNA was ber of CTCs was detected in the central compared to the peripheral detected centrally, and in sample 1, no ctDNA was detected peripher- blood sample (p = 0.04). In pairwise samples 1–6, no CTCs were ally found in the peripheral blood sample. In 12/20 ctDNA pairwise

The central blood samples were drawn on average 2 min and soluble in the blood, we speculate that ctDNA is much after compression and, according to an animal study, the less afected by physical hindrance in the capillaries as release of malignant cells starts at the manipulation proce- compared to CTCs. Hence, ctDNA blood sampling is inde- dure and stays elevated up to 60 min [19]. To the best of the pendent of blood drawing location, a fnding that could authors’ knowledge, no published study has investigated how contribute to the defnition of clinical sampling routines ctDNA levels vary with manipulation of a primary tumor in primary breast cancer for ctDNA. with regard to applied pressure. The increase of ctDNA after The major limitation of this study was the small sam- breast compression found in this study can be considered ple size and low count of CTCs, despite that a total of up relatively small, with only one case going from % MAF 0.95 to 40 ml whole blood was drawn from each patient. The to 2.36 which possibly could afect prognostication (Fig. 2d) statistical nature of CTC sampling has been described by [8]. High Ki67 were associated with increased ctDNA levels Tibbe et al. [5]. Due to the low sample size, a possible (p = 0.04) (Supplementary Table 2). diference between CTCs detection before and after breast As hypothesized, CTCs were in general signifcantly compression may have been underestimated. Similarly, the more likely to be present in central than in peripheral ctDNA analysis was limited by a relatively low plasma blood samples (p = 0.04). Six patients presented CTCs input volume, and therefore a limited number of genome only in the central samples (Fig. 3a) suggesting a dif- equivalents being analyzed for the presence of mutations. ferential CTC yield depending on sample location. The When CTCs and ctDNA markers are implemented into results are comparable to the work by Peeters et al. [23] in clinical routine, our understanding of how the concentra- metastatic breast cancer but no data from studies involving tions fuctuate during diferent interventions should be bet- diferential blood sampling of primary breast cancer are ter understood. The women in this cohort are continuously hitherto available. This diferential yield was not seen for being monitored and follow-up data will be available and ctDNA (p = 0.50). Since ctDNA is a much smaller moiety presented in future publications.

1 3 454 Breast Cancer Research and Treatment (2019) 177:447–455

Conclusion 2. Bidard FC, Michiels S, Riethdorf S et al (2018) Circulating tumor cells in breast cancer patients treated by neoadjuvant chemotherapy: a meta-analysis. J Natl Cancer Inst 110:560–567 In summary, there was no signifcant release of CTCs after 3. Bardelli A, Pantel K (2017) Liquid biopsies, what we do not know mammographic breast compression but more CTCs were (yet). Cancer Cell 31:172–179 present in central compared to peripheral blood. There was 4. Yan WT, Cui X, Chen Q et al (2017) Circulating tumor cell status a small average increase in ctDNA levels after breast com- monitors the treatment responses in breast cancer patients: a meta- analysis. Sci Rep 7:43464 pression, unlikely to be clinically relevant, and no diference 5. Tibbe AG, Miller MC, Terstappen LW (2007) Statistical consid- between central and peripheral levels was found. erations for enumeration of circulating tumor cells. Cytometry A 71:154–162 Acknowledgements The authors would like to acknowledge all patients 6. Wan JC, Massie C, Garcia-Corbacho J et al (2017) Liquid biopsies who participated in this study, the work of the SCAN-B community come of age: towards implementation of circulating tumour DNA. and the South Sweden Breast Cancer Group, and Sergii Gladchuk for Nat Rev Cancer 17:223–238 bioinformatics assistance. 7. Beddowes E, Sammut SJ, Gao M, Caldas C (2017) Predicting treatment resistance and relapse through circulating DNA. Breast 34:S31–S35 Author contributions All authors (DF, KEA, YC, AMG, CB, RR, NL, 8. Olsson E, Winter C, George A et al (2015) Serial monitoring LHS, LR) contributed to the study conception and design. Material of circulating tumor DNA in patients with primary breast can- preparation: DF, KEA, NL, LHS, LR; data collection: DF, KEA, NL, cer for detection of occult metastatic disease. EMBO Mol Med LHS, LR; sample analysis: KEA, YC, AMG, CB, RR; and statistical 7:1034–1047 analysis: DF, YC. The frst draft of the manuscript was written by DF, 9. Murtaza M, Dawson SJ, Tsui DW et al (2013) Non-invasive anal- KEA, LHS, LR, and all authors commented on previous versions of ysis of acquired resistance to cancer therapy by sequencing of the manuscript. All authors read and approved the fnal manuscript. plasma DNA. Nature 497:108–112 10. Spindler KL, Pallisgaard N, Andersen RF, Jakobsen A (2014) Funding This work was supported by the Swedish Cancer Society, Changes in mutational status during third-line treatment for meta- Swedish Research Council, VINNOVA, Mrs. Berta Kamprad Foun- static colorectal cancer–results of consecutive measurement of dation, Governmental Funding of Clinical Research within National cell free DNA, KRAS and BRAF in the plasma. Int J Cancer Health Service, Lund University Medical Faculty, Cancer Research 135:2215–2222 Foundation at the Department of Oncology Malmö University Hospi- 11. Scholer LV, Reinert T, Orntoft MW et al (2017) Clinical impli- tal, Gunnar Nilsson Cancer Foundation, BioCARE Research Program, cations of monitoring circulating tumor DNA in patients with King Gustav Vth Jubilee Foundation, and the Krapperup Foundation. colorectal cancer. Clin Cancer Res 23:5437–5445 12. Loman N, Saal LH (2016) The state of the art in prediction of Compliance with ethical standards breast cancer relapse using cell-free circulating tumor DNA liquid biopsies. Ann Transl Med 4:S68 13. Sandri MT, Zorzino L, Cassatella MC et al (2010) Changes in Conflict of interest YC, AMG, CB, RR, and LHS are shareholders and circulating tumor cell detection in patients with localized breast employees of SAGA Diagnostics AB. LHS has received honorarium cancer before and after surgery. Ann Surg Oncol 17:1539–1545 from Novartis AG. All remaining authors have declared no conficts 14. Papavasiliou P, Fisher T, Kuhn J et al (2010) Circulating tumor of interest. cells in patients undergoing surgery for hepatic metastases from colorectal cancer. Proc (Bayl Univ Med Cent) 23:11–14 Ethical approval All procedures performed were in accordance with 15. Koch M, Kienle P, Hinz U et al (2005) Detection of hematogenous the 1964 Helsinki declaration and its later amendments or comparable tumor cell dissemination predicts tumor relapse in patients under- ethical standards and ethical approval was obtained from the Regional going surgical resection of colorectal liver metastases. Ann Surg Ethical Review Board in Lund, Sweden (diary number 2014/521). The 241:199–205 experiments comply with the current laws of Sweden. 16. van Dalum G, van der Stam GJ, Tibbe AG et al (2015) Circulat- ing tumor cells before and during follow-up after breast cancer Informed consent Written informed consent was obtained from all surgery. Int J Oncol 46:407–413 individual participants included in the study. 17. Hashimoto M, Tanaka F, Yoneda K et al (2014) Significant increase in circulating tumour cells in pulmonary venous blood Open Access This article is distributed under the terms of the Crea- during surgical manipulation in patients with primary lung cancer. tive Commons Attribution 4.0 International License (http://creativeco Interact Cardiovasc Thorac Surg 18:775–783 mmons.org/licen ses/by/4.0/ ), which permits unrestricted use, distribu- 18. Martin OA, Anderson RL, Narayan K, MacManus MP (2017) tion, and reproduction in any medium, provided you give appropriate Does the mobilization of circulating tumour cells during cancer credit to the original author(s) and the source, provide a link to the therapy cause metastasis? Nat Rev Clin Oncol 14:32–44 Creative Commons license, and indicate if changes were made. 19. Juratli MA, Sarimollaoglu M, Siegel ER et al (2014) Real-time monitoring of circulating tumor cell release during tumor manipulation using in vivo photoacoustic and fuorescent fow cytometry. Head Neck 36:1207–1215 References 20. Juratli MA, Siegel ER, Nedosekin DA et al (2015) In vivo long- term monitoring of circulating tumor cells fuctuation during 1. Janni WJ, Rack B, Terstappen LW et al (2016) Pooled analysis medical interventions. PLoS ONE 10:e0137613 of the prognostic relevance of circulating tumor cells in primary 21. Nishizaki T, Matsumata T, Kanematsu T et al (1990) Surgi- breast cancer. Clin Cancer Res 22:2583–2593 cal manipulation of VX2 carcinoma in the rabbit liver evokes enhancement of metastasis. J Surg Res 49:92–97

1 3 Breast Cancer Research and Treatment (2019) 177:447–455 455

22. Förnvik D, Andersson I, Dustler M et al (2013) No evidence for 31. Saal LH, Vallon-Christersson J, Hakkinen J et al (2015) The Swe- shedding of circulating tumor cells to the peripheral venous blood den cancerome analysis network-breast (SCAN-B) initiative: a as a result of mammographic breast compression. Breast Cancer large-scale multicenter infrastructure towards implementation of Res Treat 141:187–195 breast cancer genomic analyses in the clinical routine. Genome 23. Peeters DJE, Van den Eynden GG, van Dam PJ et al (2011) Cir- Med 7:20 culating tumour cells in the central and the peripheral venous 32. Rydén L, Loman N, Larsson C et al (2018) Minimizing inequal- compartment in patients with metastatic breast cancer. Br J Cancer ity in access to precision medicine in breast cancer by real-time 104:1472–1477 population-based molecular analysis in the SCAN-B initiative. Br 24. Jiao LR, Apostolopoulos C, Jacob J et al (2009) Unique localiza- J Surg 105:e158–e168 tion of circulating tumor cells in patients with hepatic metastases. 33. Cristofanilli M, Budd GT, Ellis MJ et al (2004) Circulating tumor J Clin Oncol 27:6160–6165 cells, disease progression, and survival in metastatic breast cancer. 25. Deneve E, Riethdorf S, Ramos J et al (2013) Capture of viable N Engl J Med 351:781–791 circulating tumor cells in the liver of colorectal cancer patients. 34. Bruefer C, Vallon-Christersson J, Grabau D et al (2018) Clinical Clin Chem 59:1384–1392 value of rna sequencing-based classifers for prediction of the fve 26. Reddy RM, Murlidhar V, Zhao L et al (2016) Pulmonary venous conventional breast cancer biomarkers: a report from the popu- blood sampling signifcantly increases the yield of circulating lation-based multicenter Sweden cancerome analysis network— tumor cells in early-stage lung cancer. J Thorac Cardiovasc Surg breast initiative. JCO Precis Oncol 2:1–18 151:852–857 35. Kim D, Langmead B, Salzberg SL (2015) HISAT: a fast spliced 27. Mizuno N, Kato Y, Izumi Y et al (1998) Importance of hepatic aligner with low memory requirements. Nat Methods 12:357–360 frst-pass removal in metastasis of colon carcinoma cells. J Hepa- 36. Lai Z, Markovets A, Ahdesmaki M et al (2016) VarDict: a novel tol 28:865–877 and versatile variant caller for next-generation sequencing in can- 28. Peeters DJ, Brouwer A, Van den Eynden GG et al (2015) Circulat- cer research. Nucleic Acids Res 44:e108 ing tumour cells and lung microvascular tumour cell retention in 37. Pedersen BS, Layer RM, Quinlan AR (2016) Vcfanno: fast, fex- patients with metastatic breast and cervical cancer. Cancer Lett ible annotation of genetic variants. Genome Biol 17:118 356:872–879 29. Dawson SJ, Tsui DW, Murtaza M et al (2013) Analysis of circulat- Publisher’s Note Springer Nature remains neutral with regard to ing tumor DNA to monitor metastatic breast cancer. N Engl J Med jurisdictional claims in published maps and institutional afliations. 368:1199–1209 30. Shaw JA, Guttery DS, Hills A et al (2016) Mutation analysis of cell-free DNA and single circulating tumor cells in metastatic breast cancer patients with high CTC counts. Clin Cancer Res 23:88–96

1 3